There is very little known about methods of designing interventions for very young nonverbal children. For example, the literature and our preliminary studies suggest that many nonverbal children may acquire functional communication with an intervention program that is focused entirely on verbal communication. Other children seem to require augmentative communication systems. These augmentative systems may be effective in further producing verbal communication for some nonverbal children, and may remain as the sole effective means of communication for other children who do not acquire expressive verbal communication. The purpose of the proposed research project is to systematically compare two widely used types of intervention programs (see Koegel, 2000, NIH Autism Conference; Lord & McGee, 2001, National Research Council report; Stiebel, 1999). Both approaches are consistent with our general theoretical model in that both target the key pivotal area of motivation to communicate. The primary difference in the approaches is one accomplishes this through a primarily verbal only approach, whereas the other accomplishes this through a primarily visual approach. This difference has been of major interest to the field of autism, but has not been systematically compared in scientific research. While it does not make any difference with respect to our theoretical position as to which condition turns out to be better, it may make a large difference with respect to translation from research to practice, and with respect to understanding autism, which are major NIMH priority areas. In one condition, randomly assigned children will be provided with an intervention using a well-documented manualized intervention focused on verbal expressive communication only (Pivotal Response Training, PRT). In the other condition, randomly assigned children will receive an intervention for the same social communication functions using a well-documented manualized augmentative system of intervention (Picture Exchange Communication System, PECS). Children in the two conditions will be compared for development of verbal and nonverbal communication, changes in disruptive behavior, changes in symptoms of autism, and general adaptive behavior gains. In addition, parent satisfaction and stress measures will be gathered in order to assess the effects of each intervention on family functioning. [unreadable] [unreadable]